Download Zooplankton community structure and environmental conditions in a

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Transcript
Hydrobiologia 442: 339–350, 2001.
M. Boersma & K.H. Wiltshire (eds), Cladocera.
© 2001 Kluwer Academic Publishers. Printed in the Netherlands.
339
Zooplankton community structure and environmental conditions in a set
of interconnected ponds
Karl Cottenie, Nele Nuytten, Erik Michels & Luc De Meester
Laboratory of Aquatic Ecology, Katholieke Universiteit Leuven, De Beriotstraat 32, B-3000 Leuven, Belgium
E-mail: [email protected]
Key words: zooplankton, community structure, shallow ponds, alternative equilibria, fish predation, macrophytes
Abstract
We studied the zooplankton community structure in a set of 33 interconnected shallow ponds that are restricted
to a relatively small area (‘De Maten’, Genk, Belgium, 200 ha). As the ponds share the same water source,
geology and history, and as the ponds are interconnected (reducing chance effects of dispersal with colonisation),
differences in zooplankton community structure can be attributed to local biotic and abiotic interactions. We studied
zooplankton community, biotic (phytoplankton, macrophyte cover, fish densities, macroinvertebrate densities),
abiotic (turbidity, nutrient concentrations, pH, conductivity, iron concentration) and morphometric (depth, area,
perimeter) characteristics of the different ponds. Our results indicate that the ponds differ substantially in their
zooplankton community structure, and that these differences are strongly related to differences in trophic structure
and biotic interactions, in concordance with the theory of alternative equilibria. Ponds in the clear-water state are
characterised by large Daphnia species and species associated with the littoral zone, low chlorophyll-a concentrations, low fish densities and high macroinvertebrate densities. Ponds in the turbid-water state are characterised
by high abundances of rotifers, cyclopoid copepods and the opposite environmental conditions. Some ponds show
an intermediate pattern, with a dominance of small Daphnia species. Our results show that interconnected ponds
may differ strongly in zooplankton community composition, and that these differences are related to differences in
predation intensity (top-down) and habitat diversity (macrophyte cover).
Introduction
Several studies have investigated the relationships
between environmental variables and zooplankton
community structure in ponds or lakes situated in a
large geographic area (e.g. Pinel-Alloul et al., 1995:
area = 236 000 km2 ; Tittel et al., 1998: area =
15 300 km2 ; Keller & Conlon, 1994: area = 110 km2 ).
These studies clearly indicated a hierarchy of ecological factors. However, such large-scale surveys will
tend to reveal gross patterns determined by strong differences in water quality, or soil type, which interact
with tolerance levels of the taxa at different trophic
levels. This may impede the determination of the relative importance of specific factors that may act on
a background of local conditions. In addition, the
regional species pools that are the source of the zooplankton community of individual ponds in a large
geographic area may differ among habitats because
of the limited geographic range of certain species and
chance events associated with colonisation. Some species have even been shown to have different environmental preferences in different regions (Patalas, 1971;
Sprules, 1975). Finally, even if the ecological context is very similar, the ponds may differ strongly in
their history. Differences in fish stocking, recreational
use, inundation frequency and egg banks may all have
consequences for the actual community structure.
Part of the unexplained variance in zooplankton
community structure may stem from local community
dynamics and small-scale spatial variation. Many
studies have, for instance, indicated the importance
of biotic interactions in structuring local communities, both in relatively deep (Carpenter et al., 1985;
Carpenter & Kitchell, 1993) and shallow (Scheffer et
al., 1993; Scheffer, 1998) lakes. Scheffer et al. (1993)
developed a model of two alternative stable states in
shallow lakes. One equilibrium state, which predom-
340
inates at low nutrient concentrations, is characterised
by abundant macrophytes and clear water. This state
is stabilised by high zooplankton grazing rates, low
planktivorous and benthivorous fish abundances and
high piscivorous fish abundances. The other state is
characterised by abundant phytoplankton and turbid
water at relatively high nutrient concentrations. The
turbid state is stabilised by light limitation due to algal
blooms. At intermediate nutrient levels, both alternative stable states can occur. This illustrates that, within
a given background of abiotic conditions, biotic interactions may strongly influence both function and
community structure at different trophic levels, including the zooplankton. Irrespective of the importance of
large-scale studies, there is a need of detailed studies
on a small scale to determine how local interactions
may determine zooplankton species composition.
To evaluate the relative importance of local biotic
and abiotic interactions associated with zooplankton
community structure in a system with a similar overall ecological background, we studied a set of 33
neighbouring and interconnected ponds. The ponds
are all situated in a relatively small area (200 ha; ‘De
Maten’, Genk, Belgium), receive water from the same
sources and are interconnected, resulting in essentially the same regional species pool for the individual
ponds. Hence, it reduces the possibility of geographic
and climatic variability between the different ponds
as an explanation of differences in species composition. Moreover, the network of connecting overflows
and rivulets can easily transport individuals from one
pond to the other, so chance effects are much reduced
(Michels et al. 2001), unlike in the case of dispersal
of resting stages via waterfowl (Proctor, 1964; Proctor
& Malone, 1965). Finally, the history of the ponds is
more or less similar. All ponds were used as fish ponds
until 1990. Fish culture activities consisted of regular
carp stockings, followed by drainage (Daniëls, 1998).
Irrespective of their similar ecological background and
history, the ponds still differ widely in ecological conditions, including water transparency. As it is likely
that differences between the zooplankton communities
of the different ponds are due to differences in internal ecological processes, we relate our observations
on zooplankton communities to variation at the other
trophic levels (fish and phytoplankton densities, macrophyte cover, macroinvertebrate diversity), as well as
to abiotic characteristics (morphometric variables, turbidity, eutrophy, pH, conductivity). More specifically,
it was our aim to test the hypothesis (1) that zooplankton communities in neighbouring and interconnected
ponds can differ strongly in species composition, and
(2) that the observed differences are to an important extent related to biotic factors rather than abiotic
factors.
Study area: ‘De Maten’ (Belgium)
The ponds studied are part of the nature reserve ‘De
Maten’ (Genk, Belgium), and cover an area of 200
ha (Fig. 1). There is an altitude difference of 15 m
between the highest pond (Pond 32, NE part of the
nature reserve) and the lowest pond (Pond 1, SW part
of the nature reserve). The soil consists essentially
of sand. The marshes were converted to fish ponds
around 1400 by peat-digging and the building of dikes
(Daniëls, 1998). At present, there are 35 ponds in the
area. Fish farming stopped in 1991. The ponds are
connected with each other through a system of overflows and rivulets. The main sources of water are two
rivulets, one of which mainly feeds a subset of ponds
located in the N–W corner of the area (Ponds 18, 19,
21). The ponds are also fed by groundwater. At the
other end of the nature reserve, the outflow of water is
again diverted to the Stiemerbeek, the main rivulet.
Materials and methods
Sampling and sample processing
Quantitative zooplankton samples (cladocerans, copepods and large rotifers) were collected with a 12 l
Schindler-Patalas sampler (Vanni et al., 1997) between
8 and 10 July 1996 in 33 different ponds. Four random samples were taken in the pelagic zone of each
pond. The samples from a given pond were combined,
filtered through a 64 µm mesh net and preserved in
5% formaldehyde (final concentration). Subsamples
of 2 ml (total volume 60 ml) were taken using a
4 mm pipette (Edmondson & Winberg, 1971). All individuals present in the subsample were identified and
counted under a stereomicroscope. Different levels of
identification were used: (1) cladocerans were identified to species level because different species have
different important impacts on the ecosystem (Scheffer, 1998), (2) copepods were divided into cyclopoid
(mainly Halicyclops neglectus, Paracyclops sp., Cyclops strenuus strenuus and Acanthocyclops vennalis),
calanoid (mainly Eudiaptomus vulgaris), harpacticoid
copepods and nauplii, and (3) large rotifers were identified to the genus level. For the Cladocera, a second
341
Figure 1. Geograpical position of the study area, the nature reserve ‘De Maten’ (50◦ 570 N, 5◦ 270 E; Genk, Province of Limburg, Belgium).
Map of the pond complex showing pond numbers. Arrows indicate the in- and outflows, with the overall direction of water flow from Pond 32
to Ponds 3 and 5. The total altitudinal difference between the ponds is 15 m.
subsample of 3 ml was taken if the number of counted
individuals was less than 50 for a particular species.
For identification of the cladocerans, the following
keys were used: Flößner (1972), Flößner & Kraus
(1986) and Smirnov (1996).
Semi-quantitative samples of macroinvertebrates
were taken between 10 and 12 July 1996. The
macroinvertebrates were sampled during 8 min in each
pond using a net with 500 µm mesh size (Sutherland,
1996). This time-effort method yields reliable data
that are suitable to compare different ponds (Reid et
al., 1995). The time allocated to sample different vegetation types in each pond was proportional to the
percentage cover of the different vegetation types in
the different ponds. The animals were identified and
counted in the field. The following level of identification was used: Planorbidae, Argyroneta aquatica,
Hydracarina, Ephemeroptera, Zygoptera, Anisoptera,
Gerridae, Hydrometridae, Nepa cinerea, Ranatra linearis, Corixidae, Ilyocoris, Notonecta, Dytiscidae
(adults and larvae), other Coleoptera larvae, Hydrophilidae (adults and larvae) and Chaoborus larvae (De
Pauw & Vannevel, 1993). For each taxon and pond,
the number of individuals was assessed using a score
(0 = absent; 1 = 1 individual; 2 = 2–5; 3 = 6–10;
4 = 10–20; 5 = 20–50; 6 = > 50 individuals). We
also sampled all the ponds once during the night (July
23, 1996) using a Schindler–Patalas trap to collect
Chaoborus larvae, known to exhibit diel vertical mi-
gration behaviour in the presence of fish (La Row &
Marzolf, 1970; Berendonk & O’Brien, 1996).
Fish were sampled between 20 and 30 September 1996, because in autumn there is less sampling
stress on the animals, the young-of-the-year are included in the catches and stable population densities.
Every pond was sampled twice with a fyke (semiquantitative; catch per unit effort, CPUE) once for 2
days and once for 3 days. All individuals were counted
and identified to species level. A random sample of
40 fish (if present) were measured and the total wet
biomass for each species was determined. CPUE is
the average biomass or number of individuals per species per fyke per day. For data analysis, we lumped
the fish in three functional groups: planktivorous,
benthivorous and piscivorous fish.
Phytoplankton samples were taken at the same
time as the zooplankton samples. Phytoplankton
abundance was measured as chlorophyll-a concentration. One 250 ml sample from each lake was filtered
through glass-fiber filters (Whatman GF/C). Pigments
were extracted in methanol in the dark at 6 ◦ C for 24
h. Pigment concentration was measured spectrophotometrically and corrected for degradation products,
following the procedure of Talling & Driver (1963).
Vegetation cover was studied between 10 and 12
July 1996. We distinguished among three groups
of macrophytes: emergent vegetation (dominated by
Typha species and Phragmites australis), submerged
macrophytes (mainly Utricularia and Potamogeton
342
species) and species with floating leaves (mainly Nuphar). We used the following scores to quantify the
occurrence and abundance of the different macrophyte
types: (1) emergent species: 0 = absent, 1 = almost
none, 2 = poorly developed littoral zone, 3 = reasonably developed littoral zone, 4 = well developed
littoral zone with high abundances but low diversity,
5 = well developed littoral zone with high abundances
and high diversity. We distinguished between score 4
and 5 because it was judged that a high diversity in
emergent macrophytes creates a higher structural diversity. This problem does not occur with the other
two macrophyte types; (2) submerged macrophytes: 0
= absent, 1 = not abundant, 2 = abundant; (3) floating
vegetation: 0 = absent, 1 = less than 3% coverage, 2
= 3–5%, 3 = 5–10%, 4 = 10–20%, 5 = 20–50%, 6 =
50–100%.
Total phosphorus concentration was measured following the procedure of Murphy & Riley (1962),
adjusted to Watanabe & Olsen (1965). N–NO3
was measured semi-automatically with a Technicon
AutoAnalyzer III (Technicon corporation, 1966), after
reduction to NO2 . Fe was measured with a Perkin
Elmer 2260 Atomic Absorption Spectrometer (Perkin Elmer Corporation, 1982). Conductivity, pH and
O2 concentration were measured in situ with a DATA
SONDE 3 (Hydrolab). Water transparency was determined as Secchi disk depth (diameter 25 cm). If the
Secchi disk depth could not be measured accurately
because of the disk was still visible at the bottom of
the pond, an arbitrary Secchi disk depth of 100 cm
was used as input.
Morphometric variables of the lake were measured
in the field (depth) or using the GIS environment of
IDRISI (Version 2.0; Clark lab, 1997) after digitising
maps of the different ponds (area [module AREA] and
perimeter [module PERIM]; Michels et al., 2001).
Statistical analysis
Outliers (>3 standard deviations from the mean) in
the zooplankton data were set to values just slightly
higher than the highest value in the rest of the data set.
This ‘winsorizing’ (Sokal & Rohlf, 1981) allowed us
to work with the complete data set rather than to have
to delete outliers, and needed to be applied to approximately 3% of the data. Zooplankton species with a
total abundance less than 0.1% of the total zooplankton individuals or occurring in only one pond were not
included in the multivariate analysis. All zooplankton
densities were square-root transformed to minimize
the effect of high densities (ter Braak & Smilauer,
1998), and following the suggestion of Downing et
al. (1987) with respect to sampling variability in
zooplankton.
The general strategy in treating the environmental
variables was an a priori reduction in the number of
variables. In order to have a reliable redundancy analysis (RDA), the number of environmental variables
should be less than the number of ponds included
in the analysis (ter Braak & Smilauer, 1998). We
screened for significant correlations among variables
derived from the same trophic level, and correlated
variables were substituted by compound variables.
All the biotic variables were square-root transformed,
unless scores were used.
The macroinvertebrate data were reduced to two
compound variables: the exponent of the Shannon–
Wiener diversity index (giving the number of taxa,
but taking into account the rareness of some taxa;
Hill, 1973) and the total number of individuals in each
sample (adding the average number of individuals for
each scoring class over all taxa for each sample).
No piscivore fish species were caught in the
fykes. There was a highly significant correlation
between the planktivorous and benthivorous fish densities (both biomass and abundances) with total fish
densities (square-root transformation; biomass total
fish-planktivorous fish: r=0.82, p<0.001; biomass
total fish-benthivorous fish: r=0.97, p<0.001; densities total fish-planktivores: r=0.83, p=0.001; densities
total fish-benthivorous fish: r=0.89, p<0.001; total
densities-total biomass: r=0.91, p<0.001). We therefore used only the square-root transformed total densities as input. The ratio of square-root transformed
planktivorous fish densities on the square-root transformed benthivorous fish densities was not correlated
with total fish density, and was included to differentiate the effects of planktivory from turbidity caused by
benthivory. The three different macrophyte types were
used as independent variables.
Water transparency (Secchi disk depth), total phosphorus, N-nitrate, chlorophyll-a concentrations were
square-root transformed. pH, conductivity, oxygen
and Fe were not transformed. We used the squareroot transformed area as a measure of size of the pond.
Depth is not correlated with any of these variables, and
was used as an independent variable.
A hierarchical clustering method (Ward’s method;
Ward, 1963), based on the euclidean distances computed with standardised zooplankton densities, constructed different groups of ponds (Statistica; Stat-
343
Soft inc., 1997). Redundancy analysis (ter Braak &
Smilauer, 1998) was used to describe the relationships
between the zooplankton species (dependent matrix)
and the environmental variables (independent matrix)
(CANOCO program; ter Braak & Smilauer, 1998).
Because of interference with sampling gear (Pond 1)
and partial drainage (Ponds 4 and 5), no reliable fish
density estimates could be obtained. These ponds were
hence not included in the RDA analysis.
The automatic forward selection procedure
(CANOCO, ter Braak & Smilauer, 1998) was used to
select those environmental variables that contribute
most to the explanation of the species data set. The
automatic forward selection procedure computes the
significance of the addition of a given variable and the
stepwise cumulative variance explained with all the
selected variables in the model.
Results and discussion
Pond characteristics
The ponds studied are very shallow (mean depth of
0.5 m) and eutrophied (Table 1). Total P was in
the range of very eutrophic waters for all the ponds,
whereas chlorophyll-a concentration and secchi-disk
depths covered the spectrum from meso- to eutrophy
(Carlson, 1977). Most of the ponds had the potential
to be in both the turbid and clear-water equilibrium
state, given the threshold value of 350 µg TP l−1 set
by Jeppesen et al. (1990) for lakes smaller than 3 ha.
Zooplankton species richness and composition
The number of pelagic cladoceran species (following
Keller & Conlon, 1994: Daphnia species, Ceriodaphnia species, Diaphanosoma species, Scapholeberis mucronata, Bosmina longirostris and Chydorus
sphaericus) was comparable to the number of species
found in ponds with comparable area and depth in
other studies. In our study, an average of 5 cladoceran
species per pond was found. Dodson (1991) found an
average of 4.5 cladoceran species for ponds with an
area of less than 6 ha, and Keller & Conlon (1994)
found an average of 9 species for ponds with a depth
of less than 2 m, approximately half were cladocerans.
Within the limited range in habitat sizes of the ponds
studied, there was no relationship between the number
of species and pond area (r=−0.007, p=0.96).
The cluster analysis of the zooplankton data yielded three groups (Fig. 2). Figure 3 shows the densit-
ies of the different zooplankton taxa in the different
ponds, grouped according to the results of the cluster
analysis (Fig. 2). When the results of the clustering are
used as a second labeling variable in the results of the
RDA (Figs 4 and 5), it can be seen that there is a tendency for a gradient of Group 1, via Group 2 to Group
3. The moderate length of the species arrows and the
blurred differences between the different groups on the
graph suggest, however, that the ponds cannot be divided into clearly distinctive groups. If we combine
the results of this analysis with the original species
data (Fig. 3), some overall distinctions between the
different groups can be observed. Ponds in Group 1
are characterised by the presence of Daphnia pulex,
Polyphemus pediculus, Simocephalus vetulus and the
poor representation of Bosmina longirostris, Daphnia ambigua, cyclopoid copepods and rotifers. Ponds
in Group 2 are characterised by the presence of cyclopoid copepods, Chydorus sphaericus, Ceriodaphnia
pulchella, Daphnia ambigua, Daphnia galeata and
Bosmina longirostris. Ponds of Group 3 are mainly
characterised by a high abundance of rotifers and
cyclopoid copepods.
Alternative equilibria
Figure 4 and Table 1 shows that ponds of Group 1
were characterised by clear water, low chlorophyll-a
concentrations, low fish densities, and high diversities of macroinvertebrates. Ponds of Group 3 were
turbid, whereas ponds of Group 2 had intermediate
characteristics.
Figure 6 presents the results of the automatic selection procedure. A sharp decline in significance after
the fish densities, and the fact that half of the total
variance explained with the full model (30% compared
to 60%) is already explained by the first five selected variables (depth, turbidity, submerged vegetation,
diversity of macroinvertebrates en fish density), suggests the importance of those variables. Four of these
variables (excluding depth) are factors that play a role
in the feedback loops determining whether a shallow
water body occurs in the clear-water or turbid state
(Scheffer et al., 1993).
Substantial variation in ecological conditions was
observed among ponds, and this variation was mainly
associated with the state (clear-water or turbid state)
of the ponds studied (cf. Secchi-disk depth, phytoplankton concentration, fish abundance). It is striking
that clear-water and turbid ponds co-occurred in a system in which the ponds were so strongly connected
1
2
13
14
15
18
21
33
34
4
5
6
16
19
24
25
26
27
28
30
31
32
3
7
8
9
10
11
12
17
22
23
29
Group 1
Group 2
Group 3
7
4
6
3
3
2
4
3
7
5
4
4
2
6
9
5
7
5
6
7
3
8
7
4
4
10
7
8
7
3
1
6
6
7.4
0.3
2.6
0.6
1.6
0.5
3.0
0.4
0.5
1.3
1.7
0.4
0.5
1.1
0.4
5.2
1.3
0.7
0.2
1.4
1.9
1.3
1.8
2.8
0.3
1.3
2.9
0.6
1.4
9.5
0.3
0.2
0.1
6.6
6.4
8.6
7.2
7.3
6.8
6.0
7.7
4.9
4.3
6.3
5.7
5.2
6.3
7.8
8.1
6.2
4.8
6.4
5.8
7.5
4.4
3.7
7.9
9.15
6.8
7.8
7.5
9.1
6.1
9.6
6.9
2.6
58
34
52
73
66
21
21
37
65
60
47
17
49
11
39
34
25
48
41
49
44
43
44
37
14
34
16
39
39
36
12
48
43
7.6
7.4
7.9
7.8
7.6
7.7
7.4
7.9
7.5
7.3
7.5
7.4
7.5
7.3
7.7
7.1
7.2
7.3
6.9
7.5
8.0
7.2
7.2
7.7
5.4
7.4
7.7
7.7
8.6
6.7
4.1
7.4
6.9
1.4
0.3
0.9
4.7
4.8
3.6
6.0
0.3
1.6
1.9
2.2
3.0
1.9
0.4
0.6
0.1
1.5
1.6
1.0
1.8
0.6
0.3
0.2
0.2
0.17
1.9
1.0
2.0
0.4
0.0
0.0
1.0
0.5
60
100
32
100
25
20
25
22
30
38
37
100
100
100
56
35
53
100
100
40
100
50
67
30
100
55
100
77
100
100
100
38
100
513
550
333
277
373
263
260
273
173
200
167
397
337
550
183
187
187
167
150
183
150
173
160
437
211
550
200
197
200
200
167
200
150
488
280
289
325
226
501
506
334
271
289
316
289
235
229
248
221
231
325
361
388
411
461
361
235
198
280
266
248
235
253
221
321
339
3
4
13
5
6
30
11
14
1
3
5
1
3
11
5
4
5
1
1
4
1
3
2
20
1
2
1
5
5
15
5
3
2
3
1
1
4
2
4
5
4
3
3
5
5
3
1
2
5
4
2
2
2
4
1
1
3
4
5
4
1
5
2
4
0
0
2
0
0
0
4
0
4
0
4
0
0
1
5
0
5
0
0
5
0
0
0
0
1
5
1
1
0
0
3
6
0
1
1
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
2
0
0
2
1
2
1
0
0
0
0
0
88
9
69
1
6
8
4
95
27
13
110
57
105
10
35
26
107
32
22
9
88
95
95
5
22
62
35
60
15
93
15
24
55
3.3
1.8
3.8
1.0
2.8
1.0
1.0
3.3
6.3
2.7
7.2
4.8
6.8
3.8
1.0
5.5
4.1
3.6
4.5
2.9
3.3
5.1
3.3
1.9
4.1
3.8
7.3
7.2
5.7
4.1
5.7
3.6
7.6
2.1
2.7
1.0
24.0
1.1
21.8
0.1
3.4
0.8
0.7
2.8
—
—
5.0
2.9
0.1
0.1
5.2
0.0
1.0
7.2
19.4
8.4
0.6
—
0.0
0.6
1.6
2.0
11.7
0.0
0.7
0.0
558.7
347.5
249.5
241.7
349.6
85.7
376.3
78.7
441.5
620.7
412.9
—
—
169.4
182.2
621.2
31.8
6.2
0.2
644.9
70.7
364.9
61.3
92.8
—
690.0
6.6
14.5
22.9
50.8
0.6
159.3
5.2
Pond Depth Area O2
Cond
pH Fe
Secchi N
TP
Chla
EM
FM
SM
DenMI
DivMI PF/BF DenF
(cm) (ha) (mg l−1 ) (µS cm−1 )
(mg l−1 ) (cm)
(µg l−1 ) (µg l−1 ) (µg l−1 ) (score∗ ) (score∗ ) (score∗ ) (ind 8’−1 ) §
(CPUE)
Table 1. Environmental variables of the different ponds. The following abbreviations were used: O2 =oxygen concentration, cond=conductivity, Fe=Fe concentration, secchi=Secchi-disk depth, N=N–NO3 concentration, TP= total phosphorus concentration, Chla=chlorophyll-a concentration, EM=emergent macrophyte cover, FM=floating macrophyte
cover, SM=submerged macrophyte cover, DenMI=total density of macroinvertebrates, DivMI=diversity macroinvertebrates, PF/BF=ratio planktivorous versus benthivorous fish
densities, DenF=total density of fish. – = fish densities not available. ∗ = for the different scoring systems, see ‘Materials and methods’. §= Exponent of the Shannon–Wiener
diversity index
344
345
Figure 2. Result of hierarchical clustering analysis (Ward’s method), using the Euclidean distance of the square-root transformed ‘winsorized’
zooplankton data as dissimilarity measure. The numbering of the different clusters is obtained from the results of the multivariate analysis (Figs
4 and 5).
and shared the same water source. Timms & Moss
(1984) described a similar situation for the Norfolk
Broads in eastern England, with two ponds in a series
that were fed by the same river and approximately the
same nutrient loadings, were characterised by a very
different food web structure. Although our data on
zooplankton species composition are largely in agreement with the theory on alternative states developed
by Scheffer et al. (1993), it should be noted that our
results also suggest that there are three different types
of ponds in ‘De Maten’: clear-water and turbid ponds,
and ponds that have characteristics of both types. The
ponds in the clear-water state are characterized by
large Daphnia species (Daphnia pulex), species that
are associated with macrophytes-species (chydorids,
Simocephalus vetulus and Scapholeberis mucronata)
and species that typically occur in clear waters (Polyphemus pediculus). Although samples were taken in
the pelagic zone of the ponds, they typically also
revealed species that are associated with the littoral
zone. Timms & Moss (1984) and Irvine et al. (1989)
also observed that when macrophytes were present,
macrophyte-associated animals also occurred in the
pelagic zone of the pond. The dominance of D. pulex
and the low relative abundance of other Daphnia species suggest the absence of effective fish predation,
even though all ponds contained fish. Low fish predation leads to the dominance of large cladocerans either
through competitive superiority of large species (sizeefficiency theory, Hall et al., 1976; Gliwicz, 1990)
and/or size-selective predation by macroinvertebrates
(Gliwicz & Pijanowska, 1986; Lampert, 1987). The
ponds of Group 1 had the highest macroinvertebrate
densities and diversities of all ponds. Only in four
ponds, however, were negatively size-selective predators observed in the pelagic zone: Ponds 3 and 9
(both belong to Group 3) harboured some Leptodora
kindtii, whereas Chaoborus larvae occurred in Ponds
13 and 14. The only other predatory macroinvertebrate
species that is not strictly associated with vegetation is
Notonecta, found in Ponds 4, 13, 14, 15 and 34. Notonecta, however, is a positively size-selective predator (Murdoch & Scott, 1984). Our results therefore
suggest that the size-efficiency relationship plays a
role in determining zooplankton community structure
in the studied system. The high diversity and densities
of macroinvertebrates in ponds of Group 1 can itself
be attributed to a combination of two factors: absence
346
Figure 3. Univariate representation of the ‘winsorized’ zooplankton densities (individuals l−1 ). ‘Other chydorids’ is the sum of Pleuroxus
truncatus, Pleuroxus trigonellus, Alona quadrangula, Alonella excisa, Disparalona rostrata, Graptoleberis testudinaria. The ponds are grouped
following the results of the cluster analysis (Fig. 2).
of predation by fish and/or the presence of macrophytes, resulting in a more diverse habitat (Crowder
& Cooper, 1982; Gilinsky, 1984; Diehl, 1992; van den
Berg et al., 1997). Group 3 encompasses ponds in the
turbid state, with a dominance of rotifers (Asplancha,
Polyarthra, Brachionus and some Keratella species).
The negative relation between the presence of Daphnia and rotifers has been well documented (Fussmann,
1996). A similar difference in community structure
between clear-water ponds with large herbivores and
turbid ponds with rotifers has also been found by Reinertsen et al. (1990, 1997). The abundance of cyclopoid
copepods in the turbid ponds can be explained by the
fact that copepods are relatively efficient at escaping
fish attacks (Drenner et al., 1978).
The intermediate zooplankton community structure could be explained in two ways: (1) it could
be a transition phase between the two stable states,
or (2) a result of continuous dispersal in this highly
interconnected pond system.
A potential problem with the zooplankton data is
the fact that the ponds were sampled only once during
the seasonal succession of the different ponds. If the
seasonal succession between the different ponds is not
synchronised, it is possible that ‘different’ communities were sampled in the different ponds, especially
with respect to macrophyte development. We tried to
minimize this effect as much as possible by sampling
in July, which is the period during which macrophyte
vegetation is fully developed. Timms & Moss (1984)
observed that the macrophyte cover was quite constant
from the end of May to the beginning of August, for
three consecutive years, and it was during this period
347
Figure 4. Results of the Redundancy Analysis. Pattern of environmental (independent) variables and ponds. For abbreviations of the environmental variables, see legend Table 1. The position of the ponds is indicated by the number of the pond (see Fig. 1) and the group membership
(see Fig. 3): =Group 1, =Group 2, ×=Group 3.
that the effect of zooplankton on phytoplankton was
most noticeable.
We observed no correlation between total phosphorus and chlorophyll-a, as one would suspect in a
basically three-level trophic system (phytoplanktonzooplankton-planktivores; we observed no piscivorous
fish in ‘De Maten’). A three level trophic structure
should reduce the predation pressure of the zooplankton on the phytoplankton and let bottom-up factors
rule. The absence of the relation between total phosphorus and chlorophyll-a, however, suggests that other
forces control the densities of the phytoplankton. Macrophyte vegetation probably plays an important role
(Scheffer et al., 1993). It is, for instance, possible
that the zooplankton escapes predation by fish through
vertical or horizontal migration (Irvine et al., 1990;
Lauridsen & Lodge, 1996).
Our results indicate that neighbouring and even interconnected ponds may differ substantially in their
zooplankton community structure, and that these dif-
ferences are strongly related to differences in trophic
structure and biotic interactions. Indeed, our analysis
indicates that most of the variation in zooplankton
community structure is correlated to factors such as
fish abundance, macroinvertebrate densities and turbidity. We conclude that in the type of ponds studied
(shallow, eutrophied waters), differences in zooplankton community structure may be understood quite well
within the framework of alternative stable states developed by Scheffer et al. (1993) and Scheffer (1998).
Our results, however, indicate the possibility of a zooplankton community structure which is intermediate
to that typical of clear-water and turbid ponds.
Acknowledgements
We thank Natuurreservaten v.z.w. and especially the
warden Willy Peumans for giving us access to ‘De
Maten’ and for their cooperation. We thank the many
348
Figure 5. Results of the Redundancy Analysis. Representation of the zooplankton (dependent variables) taxa. The following abbreviations were used: A.excisa=Alonella excisa, A.quadrangula=Alona quadrangula, Asplanchna=Asplanchna, B.longirostris=Bosmina
longirostris, Brachionus=Brachionus, Calanoids=calanoid copepods, C.pulchella=Ceriodaphnia pulchella, C.sphaericus=Chydorus
sphaericus, Cyclopoids=cyclopoid copepods, D.ambigua=Daphnia ambigua, D.galeata=Daphnia galeata, D.pulex=Daphnia pulex,
D.rostrata=Disparalona rostrata, G.testudinaria=Graptoleberis testudinaria, K.cochlearis=Keratella cochlearis, K.quadrata=Keratella quadrata, Polyarthra=Polyarthra, P.pediculus=Polyphemus pediculus, P.trigonellus=Pleuroxus trigonellus, P.truncatus=Pleuroxus truncatus,
S.mucronata=Scapholeberis mucronata, S.vetulus=Simocephalus vetulus. Only the symbols representing the group a pond belongs to are shown
(see Fig. 4).
people who helped during the intensive sampling campaign, especially Eddy Holsters and Sandra Enis. We
thank Wouter Rommens for his help with the study
of macrophyte cover and Steven Declerck for help
with the study of fish communities. Nicole Podoor
measured nutrient concentrations. We thank Steven
Declerck for helpful comments. K. Cottenie is research assistant of the Fund for Scientific Research –
Flanders (Belgium) (F.W.O). E. Michels is a fellow of
the Flemish Institute for the promotion of ScientificTechnological Research in industry (I.W.T.).
This study was financially supported by project
VLINA/96/1 of the flemish government and by project GO 358.01 of the fund for Scientific Research,
Flanders.
References
Brendonk, T. U. & W. J. O’Brien, 1996. Movement response
of Chaoborus to chemicals from a predator and prey. Limnol.
Oceanogr. 41: 1829–1832.
Carlson, R. E., 1977. A trophic state index for lakes. Limnol.Oceanogr. 22: 361–369.
Carpenter, S. R., J. F. Kitchell & J. R. Hodgson, 1985. Cascading
trophic interactions and lake productivity. Bioscience 35: 634–
639
Carpenter, S. R. & J. F. Kitchell, 1993. The Trophic Cascade in
Lakes. Cambridge University Press, Cambridge.
Clark labs, 1997. IDRISI for Windows, version 2.0. Clark labs for
Carthographic Technology and Geographic Analysis, Worcester.
Crowder, L. B. & W. E Cooper, 1982. Habitat, structural complexity
and the interaction between bluegills and their prey. Ecology 63:
1801–1813.
Daniëls, L., 1998. Kansen voor natuurbehoud en -herstel. Natuurreservaten: 4–7.
349
Figure 6. Results of the automatic variable selection procedure. Both the significance of the addition of a certain variable (left Y-axis, and
solid line) as the variance explained by the selected variables (right Y-axis, and dashed line) are shown. For the list of abbreviations of the
environmental variables, see legend Table 1.
De Pauw, N. & R. Vannevel, 1993. Macroinvertebraten en waterkwaliteit. Stichting Leefmilieu, Antwerpen.
Diehl, S., 1992. Fish predation and bentic community structure:
the role of omnivory and habitat complexity. Ecology 73: 1646–
1661.
Dodson, S. I., 1991. Species richness of crustacean zooplankton in
European lakes of different sizes. Limnol. Verh. int. Ver. Theor.
Angew. 24: 1223–1229.
Downing, J. A., M. Perusse & Y. Frenette, 1987. Effect of interreplicate variance on zooplankton sampling design and data
analysis. Limnol. Oceanogr. 32: 673–680.
Drenner, R. W., R. Strickler & W. J. O’Brien, 1978. Capture probability: the role of zooplankton escape escape probability in
selective feeding of zooplanktivorous fish. J. Fish. Res. Bd Can.
35: 1370–1373.
Edmonson, W. T. & G. G. Winberg, 1971. A manual on methods
for the assessment of secondary productivity in fresh waters. IBP
Handbook Number 17. Blackwell, Oxford.
Flößner, D., 1972. Krebstiere, Crustacea; Kiemen- und Blattfüsser,
Branchiopoda; Fischläuse, Branchiura. In Die Tierwelt Deutschlands 60. Teil G. Fisher Verlag, Jena.
Flößner, D. & K. Kraus, 1986. On the taxonomy of the Daphnia
hyalina-galeata complex (Crustacea: Cladocera). Hydrobiologia
137: 97–115.
Fussman, G., 1996. The importance of crustacean zooplankton in
structuring rotifer and phytoplankton communities: an enclosure
study. J. Plankton Res. 10: 1897–1915.
Gilinsky, E., 1984. The role of fish predation and spatial heterogeneity in determining benthic community structure. Ecology 65:
455–468.
Gliwicz, M. Z. & J. Pijanowska, 1986. The role of predation in
zooplankton succession. In Sommer, U. (ed.), Plankton Ecology:
Succession of Plankton Communities. Springer Verlag, Berlin: p
253–296.
Gliwicz, M. Z., 1990. Food thresholds and body size in Cladocera.
Nature 343: 638–640.
Hall, D. J., S. T. Threlkeld, C. W. Burns & P. H. Crowley, 1974. The
size-efficiency hypothesis and the size structure of zooplankton
communities. Annu. Rev. Ecol. Syst. 7: 177–208.
Hill, M. O., 1973. Diversity and eveness: a unifying notation and its
consequences. Ecology 54: 427–432.
Irvine, K., B. Moss & H. Balls, 1989. The loss of submerged plants
with eutrophication II. Relationships between fish and zooplankton in a set of experimental ponds, and conclusions. Freshwat.
Biol. 22: 89–107.
Irvine, K., B. Moss & J. Stansfield, 1990. The potential of artificial
refugia for maintaining a community of large-bodied cladocera
against fish predation in a shallow eutrophic lake. Hydrobiologia
200/201: 379–389.
Jeppesen, E., J. P. Jensen, P. Kristensen, M. Søndergaard, E.
Mortensen, O. Sortkjaer & K. Olrik, 1990. Fish manipulations as a lake restoration tool in shallow, eutrophic, temperate
lakes 2: threshold values, long-term stability and conclusions.
Hydrobiologia 200/201: 219–228.
Keller, W. & M. Conlon, 1994. Crustacean zooplankton communities and lake morphometry in precambrian shield lakes. Can. J.
Fish. aquat. Sci. 51: 2424–2434.
Lampert, W., 1987. Predictability in lake ecosystems: the role of
biotic interactions. Ecol. Stud. 61: 323–346.
La Row, E. J. & G. R. Marzolf, 1970. Behavioral differences
between 3rd and 4th instars of Chaoborus punctipennis Say. Am.
Midl. Nat. 84: 428–436.
Lauridsen, T. L. and D. M. Lodge, 1996. Avoidance by Dapnia
magna of fish and macrophytes: chemical cues and predatormediated use of macrophyte habitat. Limnol. Oceanogr. 41:
794–798.
Michels, E., K. Cottenie, L. Neys & L. De Meester, 2001. Zooplankton on the move: first results on the quantification of dispersal
350
of zooplankton in a set of interconnected ponds. Hydrobiologia
442: 177–126.
Murdoch, W. W. & M. A. Scott, 1984. Stability and extinction of laboratory populations of zooplankton preyed on by the
backswimmer Notonecta. Ecology 65: 1231–1248.
Murphy, J. & J. P.Riley, 1962. A modified single solution method
for determination of phosphate in natural waters. Anal. Chim.
acta 27: 31–36.
Patalas, K., 1971. Crustacean plankton communities in forty-five
lakes in the Experimental Lakes Area, northwestern Ontario. J.
Fish. Res. B Can. 28: 231–244.
Perkin Elmer Corporation, 1982. Analytical methods for atomic
absorption spectrophotometry.
Pinel-Alloul, B., T. Niyonsenga & P. Legendre, 1995. Spatial and
environmental components of freshwater zooplankton structure.
Ecoscience 2: 1–19.
Proctor, V. W., 1964. Viability of crustacean eggs recoverd from
ducks. Ecology 45: 656–658.
Proctor V. W. & C. Malone, 1965. Further evidence of the passive
dispersal of small aquatic organisms via the intestinal tracts of
birds. Ecology 46: 728–729.
Reid, R. A., K. M. Somers & S. M. David, 1995. Spatial and
temporal variation in littoral-zone benthic invertebrates from
three south-central Ontario Lakes. Can. J. Fish. aquat. Sci. 52:
1406–1420.
Reinertsen, H., A. Jensen, J. L. Koksvik, A. Langeland & Y. Olsen,
1990. Effects of fish removal on the limnetic ecosystem of a
eutrophic lake. Can. J. Fish. aquat. Sci. 47: 166–173.
Reinertsen, H., J. I. Koksvik & A. Haug, 1997. Effects of fish
elimination on the phytoplankton and zooplankton in a small
eutrophic lake. Verh. int. Ver. Limnol. 26: 593–598.
Scheffer, M., 1998. Ecology of shallow lakes. Chapman & Hall,
London.
Scheffer, M., S. H. Hosper, M.-L. Meijer & B. Moss, 1993.
Alternative equilibria in shallow lakes. Trends Ecol. Evol. 8:
275–279.
Smirnov, N. N., 1996. Cladocera: the Chydorinae and Sayciinae
(Chydoridae) of the world. In Dumont, H. J. F. (ed.), Guides to
the Identification of the Microinvertebrates of the Continental
Waters of the World. SPB Academic Publishing, Amsterdam.
197 pp.
Sokal, R. R. & F. J. Rohlf, 1981. Biometry: the principles and
practice of statistics in biological research. W.H. Freeman & Co.,
New York.
Sprules, W. G., 1975. Midsummer crustacean zooplankton communities in acid-stressed lakes. J. Fish. Res. B Can. 32: 389–395.
Statsoft, inc., 1997. STATISTICA for windows (Computer program
manual). Tulsa Oklahoma.
Sutherland, W. J., 1996. Ecological sensus techniques. Cambridge
University Press, Cambridge.
Talling, J. F. & D. Driver, 1963. Some problems in the estimation
of chlorophyll-a in phytoplankton. Proceedings of a conference on primary productivity measurements. U.S. atomic energy
communitcation. TID-7633: 142–146.
Technicon Corporation, 1962. General operation instruction
manual. Tarytown, New York.
ter Braak, C. J. F. & P. Smilauer, 1998. CANOCO reference manual
and user’s guide to Canoco for Windows: software for canonical community ordination (version 4). Microcomputer Power,
Ithaca.
Timms, R. M. & B. Moss, 1984. Prevention of growth of potentially
dense phytoplankton populations by zooplankton grazing, in the
presence of zooplanktivorous fish, in a shallow wetland ecosystem. Limnol.Oceanogr. 29: 472–486. Tittel, J., B. Zippel & W.
Geller, 1998. Relationships between plankton community structure and plankton size distribution in lakes of northern Germany.
Limnol. Oceanogr. 43: 1119–1132.
van den Berg, M. S., H. Coops, R. Noordhuis, J. Van Schie & J.
Simons, 1997. Macroinvertebrate communities in relation to submerged vegetation in two Chara-dominated lakes. Hydrobiologia
342/343: 143–150.
Vanni, M. J., C. D. Layne & S. E. Arnott, 1997. ‘Top-down’
trophic interactions in lakes: effects fo fish on nutrient dynamics.
Ecology 78: 1–20.
Ward, J. H., 1963. Hierarchical grouping to optimize an objective
function. J. am. Stat. Asso. 58, 236–244.
Watanabe, F. S. & S. R. Olsen, 1965. Test of an ascorbic acid method
for determining phosphorus in water and NaHCO3 extracts from
soils. Soil Sci. Soc. am. Proc. 29: 677–678.